Family dynamics and eating disorders: the role a child is given

The family sits down. The father on the edge of the sofa, the mother closer to their daughter, the younger brother on a chair by the door. I ask how they are, and the mother answers. Then the father, briefly. The daughter speaks last, and when she finally does, all three lean forward.

In that minute I learned more than any questionnaire could tell me. Not about who is to blame — I learn nothing about that, because it is not the question. I learned how this family arranges itself around pain: who speaks, who stays quiet, who protects whom, and who is carrying all the attention right now.

Whenever I write about family dynamics and eating disorders, I have to settle one thing first, or everything else reads wrongly. Families do not cause eating disorders. That claim was part of professional lore for decades, it was harmful, and the evidence does not support it. What is true is something else, and more useful: the disorder plays out in relationships, reshapes them, and is maintained within them. Which also makes relationships one of the strongest levers for recovery.

What a role is, and why nobody hands it out

A role in a family is not something anyone assigns. It emerges from repetition. Someone calms an argument once, then again, then it is expected of them, then they feel responsible for it, and five years later that is their place in the system.

Roles are usually invisible to the very people carrying them. “That’s just how I am” is the most common description of a role acquired at the age of seven. And roles complement each other: if one person carries the worry, another carries none; if one expresses anger, another may not; if one is the problem, another gets to be fine.

In families where an eating disorder appears, I often meet a child whose role until then was “the one we never worry about”. Excellent at school, considerate, adaptable, quiet. That is not a coincidence but a logical consequence: a child who may not make room for their own needs will express them in a way that does not look like a demand.

The child who carries the symptom

In family therapy we use the term identified patient. It names the member who carries the symptom and whom the family brings in: “something is wrong with her, please help her”.

The term is not there to minimise the pain or spread it across everyone. The illness is real and dangerous, and one person has it. It says something else: that the symptom also performs a task nobody intended. While it lasts, the parents are occupied with it rather than with what lies between them. A family that was drifting apart has a shared task. A child who could never ask for anything has everyone’s attention — at a price wildly out of proportion to what they gained.

This is the most uncomfortable part of this piece, so let me be plain: none of it is deliberate. Nobody invented it and nobody lives it consciously. Which is why guilt is not a useful category here — I have written about the difference between guilt that moves you and guilt that paralyses you in the piece on healthy and toxic guilt.

Triangulation: when a child becomes a bridge

Tension between two adults is rarely contained between them. As a rule it is routed through a third person. In family therapy we call this triangulation, and it is one of the most reliable patterns I see.

It looks innocent. A mother tells her daughter what bothers her about the father, because “there’s no talking to him”. A father asks his daughter not to mention that he stayed late at work again. The child becomes translator, comforter and judge all at once — a role far too large, and one they have no tools for.

A child in that position learns to read adults’ moods faster than their own. Later everyone praises the resulting skill: empathic, mature, perceptive. Underneath it is a person who does not know what they need, because they never had anywhere to practise. When enough of their own needs accumulate without language, food is not a bad candidate to become the arena where all this gets expressed.

Parentification and “the good one”

The same pattern continues as parentification: a child who takes over part of the parental role. Looking after the younger ones, holding the house together, watching a parent and assessing how much they can bear today.

This is not always the result of neglect. It often happens in very close families that have been hit by something — illness, loss, a parent’s long distress, a move, a separation, addiction. The child responds with what they know: they become useful.

The price is precise. Such a child develops an outstanding capacity to adapt and a very weak capacity to ask for anything for themselves. And because adaptability was always praised, their first signs of distress do not look like rebellion but like even greater discipline. Which is why they are so often missed. If this describes you, the piece on setting boundaries is worth reading.

Why the system resists when the child starts moving

Families, like any system, tend to preserve the balance they have. That is not a bad thing — it is what holds a family together in a crisis. It becomes a problem when the balance is built around a symptom.

So in therapy I often see this: the young person starts to move, and something in the family gets worse. The parents start arguing again. A sibling who was never any trouble suddenly is. Someone falls ill. This is not sabotage and not a sign that things are going wrong. It is a sign that the tension was being held somewhere and is now looking for a new place.

That moment is at once the most fragile and the most fertile point of the work. If the family recognises it as an expected part of the process, they can move on. If they read it as proof that “things were better before”, the system closes back up.

What Minuchin found, and what was made of it

Now a section that argues against much of what is still said on this subject — including things that were long part of my own field.

In the 1970s Salvador Minuchin described the so-called psychosomatic family: enmeshment, overprotection, rigidity and unspoken conflict. His clinical descriptions are sharp, and many families recognise themselves in them. The problem lies in the conclusion others drew: that such families cause the disorder.

That conclusion does not hold. The observations were made in families who had already lived for months or years with a seriously ill child — and enmeshment and overprotection are entirely understandable responses to the fear that your child is dying. Cause cannot be separated from consequence there. There is no prospective evidence that a particular family style predicts the onset of an eating disorder.

The cost of that error was not theoretical. A generation of parents were told they were the reason, some treatments deliberately separated them from their child, and many families avoided help as a result. The Academy for Eating Disorders has since published a position paper explicitly rejecting it. Family factors exist, but they are one among many and never the only one.

So what does work

The irony is that the best results arrived precisely when the field stopped working on blame and started working on competence.

Family-based treatment, in the Maudsley tradition, puts parents into an active role. Not as a cause to be corrected, but as a resource available to the young person every day that no therapist can replace. For adolescents with anorexia, guidelines make it the first choice, not the last stop.

What matters is what this does to the roles. Parents stop being spectators and get back the place that belongs to them. The child stops being the sole carrier of the problem. The illness is named as something separate from the young person — which lets parents be firm with the illness and gentle with their child at the same time. Adults are treated differently, but the principle holds: nobody recovers alone, in an empty room.

Where this view does not hold either

And now against myself. The systemic view is my native language, so I have to say where it fails.

First, there is not something to find in every family. People sit with me who acted well, had a close relationship and clear boundaries, and their child has an eating disorder. Heritability in these conditions is substantial, and so is temperament. If I insist on hunting for a pattern in such a case, I will invent one — and hand the family a burden that is not theirs.

Second, the systemic view must never cover violence. Where there is abuse or neglect, talk of “circles and arrangements” obscures a responsibility that runs one way only. There the pattern is not neutral and I do not treat it as such.

Third, and most practically: where health is at risk, family dynamics are not the first address. With fainting, heart rhythm disturbances, loss of periods, rapid deterioration or self-harm, the young person sees a doctor first. Relationship work alongside an unstable physical state is not an ethical choice. I have written about when a doctor comes before everything in the piece on when a doctor is the first step.

What changes when a child puts the role down

The best part of this work is not the moment the symptom eases. It is the moment someone in the family first does something their role never allowed.

When the father who was for years “the one who isn’t there” is the first to notice and say he is worried. When the mother stops explaining her daughter’s feelings and waits for her to say them. When the sibling says how much they minded that nobody ever asked them anything. When the daughter says she does not want something — and nobody falls apart.

These small things shift more than any explanation. A role does not disappear because you recognise it. It disappears when you step out of it once and survive.

What you can do as a parent

You do not need to analyse your family to begin. A few concrete moves are enough.

Take your concern directly to the person it concerns, not through your child. Ask all your children how they are, not only the one in trouble — siblings in these families tend to go quiet. Take responsibility at mealtimes when that is needed, without negotiation but also without explaining to your child what they think about themselves. Find your own space, because an exhausted parent is available to nobody. And come for help even if your child refuses — parents may come alone.

How to choose the person you will work with I described in the piece on choosing a therapist.

What to remember

Families do not cause eating disorders. That was an error of the profession that cost many people years, and it is not worth repeating.

What is true is subtler: in every family roles arrange themselves, and when a symptom appears the system arranges itself around it. The child who carries the symptom is often also maintaining something with it — peace between the parents, silence about something else, the appearance that everything is fine. Nobody chose this and nobody is to blame.

And last: precisely because the disorder plays out in relationships, relationships are the strongest tool we have. Parents are not a problem to be removed from the room. Most of the time they are what makes movement possible at all. I have written more about what a symptom says within a system in the piece on the symptom as a message.